Interpretation of Factor Loadings in CFA

The factor loadings in a CFA estimate the direct effects of unobservable constructs on their indica- tors.If an unstandardized factor loading is 2.0 for the direct effect of Customer Delight ® Delight1, then we expect a two-point difference in the indicator Delight1 given a difference of 1 point on the factor of Customer Delight.While unstandardized estimates can be insightful, they are rarely reported in the results of a CFA. Standardized estimates are most frequently reported because they allow you to compare the weights of indicators across a CFA. Standardizing an estimate converts your factor loading to a 0 to 1 scale, which allows for an easier comparison of indica- tors. Additionally, squaring the standardized factor loading will give the proportion of explained variance (R2) with each indicator.This lets you know how much of the variance in the indicator is explained by the unobserved construct. For instance, if a standardized factor loading is .80, then the unobserved variable explains .802 = .64, or 64% of the variance of the indicator.

How do I know if I have an acceptable indicator? If you have a standardized factor loading that is greater than .70 or explains at least half of the variance in the indicator (.702 =.50), then your indicator is providing value in explaining the unobserved construct. If you are not explaining at least half of the variance, that indicator is contributing little to the understand- ing of the unobservable construct. Once we have determined the standardized value for each factor loading, we can also determine the measurement error for each indicator.The measure- ment error for each indicator is simply 1 − R2.Thus, the lower the explained variance is in an indicator, the higher the measurement error.

Source: Thakkar, J.J. (2020). “Procedural Steps in Structural Equation Modelling”. In: Structural Equation Modelling. Studies in Systems, Decision and Control, vol 285. Springer, Singapore.